Abstract B1-11: Application of the CN-AVG method to reconstruct the evolutionary history of glioblastoma multiforme

Abstract

Abstract We have constructed evolutionary histories of cancer genomes using a new method, the Copy Number Ancestral Variation Graph (CN-AVG). Unlike previous cancer evolutionary studies which infer the clonal structure and evolutionary history of a tumor using single nucleotide variant frequencies, the CN-AVG method predicts the ordering of structural rearrangements, such as inversions, duplications, and deletions, using copy number changes and breakpoints derived from whole genome sequencing data. Because finding the most parsimonious ordering of genomic rearrangements is an NP hard problem, we build a consensus history by merging a large number of possible evolutionary histories generated from MCMC sampling. We tested the CN-AVG method on simulated rearranged genomes and achieved an average accuracy of 62%. We also found that accuracy decreased with increasing complexity of the simulated rearrangements, as expected. We then applied the method to TCGA Glioblastoma Multiforme genomes to search for evolutionary patterns in this disease. The CN-AVG method may distinguish driver mutations as the early rearrangement events and passenger or secondary mutations as the later events in the reconstructed evolutionary history of the tumor, and will be powerful for both the clinic and for research. Citation Format: Tracy J. Ballinger, Daniel Zerbino, Benedict Paten, David Haussler. Application of the CN-AVG method to reconstruct the evolutionary history of glioblastoma multiforme. [abstract]. In: Proceedings of the AACR Special Conference on Computational and Systems Biology of Cancer; Feb 8-11 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 2):Abstract nr B1-11.